Abstract Archives of the RSNA, 2010
SSM19-06
Reliability of Renal Length Measurements on Ultrasound Compared to Multiplanar CT
Scientific Formal (Paper) Presentations
Presented on December 1, 2010
Presented as part of SSM19: Pediatrics (Genitourinary)
Mariana Laura Meyers MD, Abstract Co-Author: Nothing to Disclose
David B. Larson MD, Presenter: Nothing to Disclose
Sara Marie O'Hara MD, Abstract Co-Author: Author, Amirsys, Inc, Salt Lake City, UT
Stockholder, Amirsys, Inc, Salt Lake City, UT
Speakers Bureau, Toshiba Corporation
Sonographic renal length measurements are often used to evaluate renal growth in children at risk for renal scarring. There is little information in the literature comparing ultrasound renal length measurements to a gold standard. We performed a retrospective study of patients imaged with ultrasound and CT to: 1) establish CT multiplanar reformats measurements as the gold standard 2) to assess the accuracy of ultrasound in measuring the renal length, and 3) to compare the precision of US renal length measurements with expected annual renal growth.
Seventy six patients who underwent renal or abdominal ultrasound and abdominal CT within two weeks of each other were evaluated. All sonographic renal length measurements documented on the images were included in the study. Scanning positions were recorded. Two of the authors measured the renal length using 3 mm coronal multiplanar reformatted CT images in separate occasions and the interobserver and intraobserver variability were calculated. Sonographic renal length measurements were compared to those obtained by CT, for accuracy and precision. The statistical analysis also included comparison of variability of sonographic renal measurements to expected renal growth.
The mean absolute value of interobserver error was 0.9 mm, with standard deviation of 0.8 mm. The intraobserver error variability for both readers was less than 1 mm. Compared to CT, single sonographic measurements underestimated renal length by a mean of 1.46 mm, with standard deviation of 5.61mm (p <0.01). Sonographic measurements with patients in decubitus position were slightly more accurate than supine and prone positions and were less precise for supine than decubitus and prone positions. The 95% confidence interval of sonographic measurement corresponded was +- 11 mm, corresponding to approximately 4 years of renal growth.
Renal length measured on multiplanar CT reformatted images are highly reproducible between observers and can be used as a gold standard. Ultrasound renal length measurements slightly underestimate renal length as measured by CT. Precision of sonographic renal measurement is limited, with the confidence interval corresponding to several years worth of renal growth.
The frequency with which serial renal ultrasounds should be performed to evaluate renal growth is limited by the precision of ultrasound measurements.
Meyers, M,
Larson, D,
O'Hara, S,
Reliability of Renal Length Measurements on Ultrasound Compared to Multiplanar CT. Radiological Society of North America 2010 Scientific Assembly and Annual Meeting, November 28 - December 3, 2010 ,Chicago IL.
http://archive.rsna.org/2010/9004080.html